Phase Splitter: Impedance Puzzling

[Warren]

I'm looking at the phase splitter design, discussed briefly in the Art of
Electronics, Fig 2.27 (in an old edition, later chapters in the new one):

https://picasaweb.google.com/lh/photo/YpOlkMD5Q9ZEAvWPDvhQRA?
feat=directlink

( http://tinyurl.com/3g8uxee )

I'm rather puzzled by the LTspice results from the DC transfer function
(with coupling cap shorted, and DC offset set to operating Vb):

--- Transfer Function ---

Transfer_function: 0.993969 transfer
v2#Input_impedance: 39651.7 impedance
output_impedance_at_V(out0): 27.2453 impedance

Rin seems about right. But Rout0 & Rout180 seem way off.

Isn't Zout (at emitter) = Zsource / ( Beta + 1 ) ?

But if I put a 50 ohm Rs in the circuit where the short circuit is shown
(and unhook C1), the Rout0 only increases very slightly from 27.5 ohms.
With this calculation, Zsource would have to be about 8.3k to arrive at
27 ohms, since Beta=300. Or is Zsource something else here?

The other puzzling thing is why is Rout180 reading about 5.6k ohms? With
R1=4.7k ohms, I have trouble with that answer.

LTspice file below.
Thanks, Warren

..............................................
Version 4
SHEET 1 880 680
WIRE 208 -112 32 -112
WIRE 528 -112 208 -112
WIRE 32 -80 32 -112
WIRE 208 -80 208 -112
WIRE 528 -80 528 -112
WIRE 32 16 32 0
WIRE 32 16 -128 16
WIRE 208 32 208 0
WIRE 272 32 208 32
WIRE 352 32 272 32
WIRE 208 48 208 32
WIRE -288 96 -304 96
WIRE -128 96 -128 16
WIRE -128 96 -288 96
WIRE -80 96 -128 96
WIRE 32 96 32 16
WIRE 32 96 -16 96
WIRE 96 96 32 96
WIRE 144 96 96 96
WIRE -288 176 -288 96
WIRE 208 176 208 144
WIRE 272 176 208 176
WIRE 352 176 272 176
WIRE 32 192 32 96
WIRE 208 192 208 176
WIRE -288 304 -288 256
WIRE 32 304 32 272
WIRE 32 304 -288 304
WIRE 208 304 208 272
WIRE 208 304 32 304
WIRE 528 304 528 0
WIRE 528 304 208 304
WIRE 208 336 208 304
FLAG 208 336 0
FLAG -304 96 In
IOPIN -304 96 In
FLAG 352 32 Out180
IOPIN 352 32 Out
FLAG 352 176 Out0
IOPIN 352 176 Out
DATAFLAG 96 96 ""
DATAFLAG 272 32 ""
DATAFLAG 272 176 ""
SYMBOL Misc\\battery 528 -96 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V1
SYMATTR Value 20V
SYMBOL npn 144 48 R0
SYMATTR InstName Q1
SYMATTR Value 2N3904
SYMBOL res 192 -96 R0
SYMATTR InstName R1
SYMATTR Value 4.7k
SYMBOL res 192 176 R0
SYMATTR InstName R2
SYMATTR Value 4.7k
SYMBOL res 16 -96 R0
SYMATTR InstName R3
SYMATTR Value 150k
SYMBOL res 16 176 R0
SYMATTR InstName R4
SYMATTR Value 56k
SYMBOL cap -80 112 R270
WINDOW 0 32 32 VTop 0
WINDOW 3 0 32 VBottom 0
SYMATTR InstName C1
SYMATTR Value 4.7uF
SYMBOL voltage -288 160 R0
WINDOW 123 0 0 Left 0
WINDOW 39 24 132 Left 0
SYMATTR InstName V2
SYMATTR Value SINE(5.3134 .5 60Hz)
TEXT -288 360 Left 0 !;tran 0 .2 0 .001
TEXT 344 0 Left 0 ;5597ohms
TEXT 352 144 Left 0 ;27.5ohms
TEXT -88 360 Left 0 !.tf V(Out0) V2
TEXT -296 -8 Left 0 ;Shorted for Transfer\nFunction only..

 

[holyhigh]

I'm looking at the phase splitter design, discussed briefly in the Art of
Electronics, Fig 2.27 (in an old edition, later chapters in the new one):

https://picasaweb.google.com/lh/photo/YpOlkMD5Q9ZEAvWPDvhQRA?
feat=directlink

( http://tinyurl.com/3g8uxee )

I'm rather puzzled by the LTspice results from the DC transfer function
(with coupling cap shorted, and DC offset set to operating Vb):

--- Transfer Function ---

Transfer_function: 0.993969 transfer
v2#Input_impedance: 39651.7 impedance
output_impedance_at_V(out0): 27.2453 impedance

Rin seems about right. But Rout0 & Rout180 seem way off.

Isn't Zout (at emitter) = Zsource / ( Beta + 1 ) ?

But if I put a 50 ohm Rs in the circuit where the short circuit is shown
(and unhook C1), the Rout0 only increases very slightly from 27.5 ohms.
With this calculation, Zsource would have to be about 8.3k to arrive at
27 ohms, since Beta=300. Or is Zsource something else here?

The other puzzling thing is why is Rout180 reading about 5.6k ohms? With
R1=4.7k ohms, I have trouble with that answer.

LTspice file below.
Thanks, Warren

agree with you on the Out180, it is close to 4.7k only 10% high (would
think it would be lower)

agree with you on the Out0 too, should be near ((150k ll 56k) / hfe ) ll
4.7k => perhaps 1k ?

 

[Warren]

Warren expounded in

I'm looking at the phase splitter design, discussed briefly in the
Art of Electronics, Fig 2.27 (in an old edition, later chapters in
the new one):

https://picasaweb.google.com/lh/photo/YpOlkMD5Q9ZEAvWPDvhQRA?
feat=directlink

( http://tinyurl.com/3g8uxee )

...

Ok, I've figured this out now-- it is technically:

Zout = ( re + Zsource / ( Beta + 1 ) ) || R2

which is pretty much just re||R2, which is still just re.
This makes sense to me now.

The problem with Rout180 now seems like it was
pilot error- when I tried it again, I got the
answer 4698.3 ohms, which is sensible. Pretty
much just Rc as expected. I must have forgot to
change some V2 parameters the first time around
for the .tf function.

So nevermind--

"Back to your gas chamber, ahem, chamber music.."

Warren

 

[Warren]

Jim Thompson expounded in

A better way to do the "transfer function": Change V2 to a current
source and sweep positive and negative. Plot output voltages versus
voltage at base of Q1.
...Jim Thompson

Is this what you had in mind (DC sweep)?

https://picasaweb.google.com/lh/photo/J66Sm9JWl6RnBSao9S6k9w?
feat=directlink

This is ok for showing response(s) to signal at Vb. I do like the fact I
don't have to worry about the coupling cap. But one thing I still like
about the .tf output in LTspice is the impedance measurements. The .tf
command only works with a voltage source.

Warren